FX Daily: Clues versus evidence

NFIB and ISM surveys have offered clues that US employment and inflation are about to slow down. Today, we’ll see whether US payrolls declined below 200k in March, which we think is the benchmark for the market’s reaction. There are risks of another leg lower in the dollar, and the Canadian dollar is facing a double threat as domestic jobs data is released too.

A couple of data releases this week have hinted that the strong US activity story may be about to turn. The ISM services index declined more than expected, but we were mostly interested in the “prices paid” component, which slowed meaningfully to a four-year low. Yesterday, the NFIB reported only 11% of small businesses are looking to create jobs in the next three months, the lowest percentage since May 2020. Small businesses account for roughly half of total US jobs and the hiring indicator has proven to have better predictive power (with a three-month delay) of official payrolls than other series in the past year and a half. At present, the March NFIB and ISM surveys – both services and manufacturing – suggest we could see sub-50k payrolls by June.

Today, the March payrolls are expected at 214k and our US economist is calling for 200k. We think a print below 200k should put pressure on the dollar, endorsing the recent signs that the employment story is softening and that the Fed will be in a comfortable position to start cutting in the summer. Expect a big market impact from revisions: last month, the December and January prints were revised lower by a total of 165k.

The very high sensitivity of the dollar and the FX market to US data is unlikely to fade, and we doubt markets will feel particularly restrained to price in more Fed cuts/add dollar shorts even if the turn in activity and inflation figures prove only gradual. A May cut looks off the table, but June remains a tangible possibility. Should the pricing for a June cut move from 60% to 100%, the dollar may well take a bigger hit than what the swing in rate differentials would imply, as markets accelerate the rotation to pro-cyclical currencies, and in particular those backed by central banks in no rush to cut. We continue to like the Norwegian dollar in the G10 space, as well as the antipodeans, while the Canadian dollar should be a laggard.

The revision higher in the eurozone’s March PMIs yesterday meant the composite index no longer points to contraction, mostly driven by an improved outlook for services. That has helped EUR/USD consolidate above 1.0800, even though the primary driver of the pair remains US data. A sub-200k US payrolls print today should pave the way for a re-test of the 1.0875 100-day resistance.

As discussed in our European Central Bank preview, recent data releases have had something for both the doves (lower inflation) and the hawks (better activity indicators). That may warrant some lingering caution and reiteration of data dependency by President Christine Lagarde next week, although the overall message may sound increasingly dovish and hint more clearly at a June cut. That is what the March minutes seemed to suggest too. It’s also worth reminding how Lagarde unexpectedly switched to aggressive guidance when she signalled a hike back in June 2022: will the doves convince her to commit to a cut this time? Another European central bank – the Riksbank – did so recently, signalling easing by the end of the second quarter. Such explicit guidance by the ECB is not our base case for next week, but one risk not to be dismissed.

February retail sales data will be released in the eurozone today: expectations are for a month-on-month decline, and modest improvement in the depressed year-on-year figure, but the market impact should be negligible.

Canadian jobs data will be released at the same time as the US release today. Consensus expectations are 25k, with almost all survey responders expecting a softer print from the strong 40.7k February reading.

The Canadian dollar has often shown a positive correlation – or less negative correlation compared to other high-beta peers – to key US data releases of late. That is because Bank of Canada rate expectations remain very strictly tied to the Fed’s and, at times, almost insensitive to Canadian data.

After a strong end of the quarter, CAD is trading on the soft side against almost all G10 currencies this week, and today’s releases are a key risk event. A combination of sub-consensus payroll prints in the US and Canada could send the Canadian dollar meaningfully lower against G10 peers – especially other commodity currencies. Next week, the Bank of Canada meeting could see policymakers open the door to a June cut, and we think the market pricing for 75bp of easing by year-end may be too conservative. 

This morning we saw retail sales in Romania and industrial production and retail sales in Hungary. Later, retail sales in the Czech Republic will also be released. However, the highlight of today will be the National Bank of Poland press conference and the governor’s comments on yesterday’s decision to leave rates unchanged as expected. The focus should be on the inflation outlook in the second half of the year despite the downside surprise in inflation these months. Government measures and their adjustments will be key here. As we mentioned yesterday, we expect the press conference to be more hawkish despite spot inflation, which should be positive for the zloty.

Given the improvement in global sentiment and despite the repricing in PLN rates, the still high interest rate differential has already moved EUR/PLN closer to 4.280 yesterday, which is our level for these days. However, we believe the press conference still has the potential to support FX and reach this level today. In Romania, yesterday the central bank also left rates unchanged as expected and the press release was rather neutral, if anything. For now, we see no signs of a coming rate cut however we believe this is the next step for the May meeting unless we see a major upside surprise in inflation next week.

Read the original analysis: FX Daily: Clues versus evidence